This is a revised proposal to validate novel biomarkers of alcohol use, and to use the validated biomarkers to examine the validity of new and standard self-reported measures of alcohol use, among persons infected with HIV in Uganda. Low cost and free HIV antiretroviral therapy (ART) has been increasing in sub-Saharan Africa. However, alcohol use is pervasive in Uganda and may threaten treatment outcome by impairing treatment adherence, increasing the risk of medication related hepatotoxicity, and reducing the bioavailability of ART, leading to sub-therapeutic medication levels and viral resistance. Measuring alcohol exposure in Uganda is problematic because (1) ubiquitous home-brews are served in non-standard drink volumes, (2) the alcohol concentration in these may vary widely, (3) the identification of alcohol related problems may be culturally dependent, and (4) persons receiving or hoping to receive ART that is in short supply may under- report their alcohol consumption. Several biomarkers of short-term and prolonged heavy alcohol use (ethyl glucuronide, ethyl sulfate, phosphatidyl ethanol, and fatty acid ethyl esters) have shown outstanding test characteristics in North America and Europe but have not been tested in Africa. In this study, 75 HIV- infected persons in Mbarara, Uganda, will be recruited from an existing NIH-funded cohort of HIV infected persons, and a high seroprevalence hospital-based HIV counseling and testing program. We will stratify by gender, ART status, and recent drinking, including equal numbers of abstainers, moderate drinkers, and heavy drinkers. We will first conduct baseline interviews using newly developed alcohol consumption and standard alcohol screening questionnaires and collect specimen samples. Next, we will validate the biomarkers using breathalyzer tests conducted at daily home visits, over a period of three weeks. Last, we will use biomarkers with high validity to examine the utility of the new and standard alcohol consumption surveys administered at baseline. PUBLIC HEALTH SUMMARY: Alcohol use is pervasive but difficult to measure in Uganda, as most drinks are home-brews, and served in non-standard drink sizes, with varying alcohol concentration. Alcohol use may impact HIV treatment adherence and HIV outcomes, therefore valid measurements of alcohol use are needed to assess HIV treatment success. This study will validate novel biomarkers and new and standard self-reported measures of alcohol use among HIV positives in Uganda. [unreadable] [unreadable] [unreadable] [unreadable]